Discover Life in America

John Longino - 21 July, 1999

barcodes, 21 July

Date: Wed, 21 Jul 1999 09:01:45 -0800
To: barcode group:;;@elwha.evergreen.edu;
From: "John T. Longino" <longinoj@evergreen.edu>
Subject: barcodes, 21 July

In response to Steve's of 21 July:

Well, Steve is forcing me to consider having the symbology contain
only part of the unique identifier. I could be convinced that the
advantages outweigh the costs.

First I have a question for Steve: as I understand it from your
message, your label looks like

|||||||||||||||
SM1234567
KUNHM-ENT

So what exactly is the unique identifier in the database?
SM1234567KUNHM-ENT? SM1234567 KUNHM-ENT? KUNHM-ENT SM1234567?

It seems to me there should be nothing else on the label except the
symbology and the unique code, exactly as it should appear in a
database. There should be no ambiguity about exactly which part of
the human readable label is the identifier. Also for this reason I
don't like spaces in unique codes. They make me nervous. Brian Brown
at LACM uses spaces ("LACM ENT 123456") and I kind of wish he didn't.

What are some consequences of having the symbology contain only part
of the code? In a mixed collection it means the operator will not
know what part of the code is contained in the symbology. In some
cases it might be the whole code, in others a portion of the code.
This means whenever a scanner is used for data entry, the code or
list of codes will need to go into some kind of buffer where the
operator can look at them and add the missing parts of the code.

I use Colwell's Biota to manage both the ALAS database and my own
database on ants. It is designed for the opposite situation, where
the symbology contains the full code, and you want to recognize a
particular prefix for various reasons. At this point it would lose a
lot of functionality if it had to deal with partial codes.

Partial codes might tempt people to use partial codes in their
databases, with the idea that the prefix can be added later or when
the data are exported. Project ALAS did this up until recently. The
unique identifiers on the labels are of the form INBIOCRI001234567,
and what we stored in the database was IB1234567. We decided this was
a dangerous precedent, and would cause problems later. I was already
having problems keeping my ant database coordinated with the ALAS
database, because I used full prefixes in the ant database (where I
also have LACM codes and Univ. Oklahoma codes, hence the concern
about mixed codes). It was also a pain remembering to add the full
prefixes when exporting the data to collaborators. The original
reason for the abbreviation was to save space, but saving disk space
seems increasingly irrelevant in today's world. We have just
converted the ALAS database to have full codes. This will be
increasingly important as databases are made accessible on the Web,
and global searches for particular specimen codes will be possible.

If we do accept partial symbology, what software tools will be
needed? Your institution's prefix would need to be added
automatically (because most of the material you handled would have
your institution's prefix), unless you override it by choosing some
other prefix, perhaps from a pop-up list. Steve, how do you handle
this at Snow?

Enough for now. Steve, thanks for adding so much to this discussion.
I hope you are willing to keep it up!

Jack
******************************************************
John T. Longino
Lab I, The Evergreen State College
Olympia WA 98505 USA
longinoj@evergreen.edu
Ants of Costa Rica on the Web at http://www.evergreen.edu/ants
Project ALAS at http://viceroy.eeb.uconn.edu/ALAS/ALAS.html
******************************************************






Discover Life in America | Science | Unique Identifiers & Barcodes | Correspondence | John Longino - 21 July, 1999